RCD FCU tripped when fitted to CH, likely reasons?

[ericmark]

The central heating did run on a RCBO until solar panels fitted, no problems, when solar panels fitted it was transferred to a UPS supply, so no longer fed through the RCBO, so I fitted a RCD FCU, but when CH started it tripped, so put old FCU back, thought summer is time to fiddle, not winter.

So out of interest measured the current through the earth cable, and no where near 15 mA, there is some current around 5 mA, but not enough to trip a RCD.

Same measuring meter tails, earth leakage is well within limits. Not used an insulation tester yet, that's a summer job, but it does seem something to do with using a RCD FCU instead of a RCBO.

I could not get a type A RCD FCU from City Electricial factors, so wonder if just unlucky, my RCD tested died, so can't test the RCD without borrowing one from work, have considered setting up a test so can simple plug in to test and take it to work to test, here no socket outlets without RCD protection. There are at work.

But does seem like a faulty RCD FCU. But wondered if anyone else has had problems with RCD FCUs as to date found very few faulty RCD's

 

[DetlefSchmitz]

I'm not sure why you weren't measuring the difference in current between the L and N, like the RCD would be doing.

 

[ericmark]

I'm not sure why you weren't measuring the difference in current between the L and N, like the RCD would be doing.

I did and 8 mA is not enough to trip a 30 mA trip, I could not reach the two lives at the FCU, that is the leakage to whole of the solar panel system and inverter.

I wanted some pictures of using the clamp on, and has not intended to look at the central heating, I had decided last year to wait until will did not need heating before investigating further, after all it all complies as it is, and I have a compliance certificate to prove it. However it niggles me to have the central heating as the only item in the house not RCD protected.

I did note a leakage to the earth rod and I was wondering if over time it could cause erosion, but 7 and 8 mA is really nothing, I had expected to find the inverter had a much higher leakage and that there was a fault with the central heating some where which was causing the RCD to trip.

I want to add a second thermostat, and thought it would be good to cure the leakage problem first, but can't really cure a problem which is not there.

 

[ebee]

here no socket outlets without RCD protection

Hiya Eric.
You could always do the updown RCD test thingy.
Make up an extension lead (guard it with your life - not to be used by anyone else) . At plug end omit E. At socket end L to L and both N E to N. A sort of "Fake TNC-S" extension lead that can be plugged in to RCD protected sockets and will not trip your house RCD but capable of testing RCD units at the extension socket end . A bit naughty but a handy way to run RCD tests from mains power on an RCD board.

Actually perhaps I should have said, wire plug normally ad at socket end park up the E conductor in a connector block (Wago perhaps) and use the L & N normally but put a link in to connect the N & E terminals , therefore your E (on the socket) is not an Earth but really a Neutral !

 

[davelx]

What type of UPS? True sine-wave?

 

[Murdochcat]

Does it require RCD protection ?

 

[AdrianUK]

Hiya Eric.
You could always do the updown RCD test thingy.
Make up an extension lead (guard it with your life - not to be used by anyone else) . At plug end omit E. At socket end L to L and both N E to N. A sort of "Fake TNC-S" extension lead that can be plugged in to RCD protected sockets and will not trip your house RCD but capable of testing RCD units at the extension socket end . A bit naughty but a handy way to run RCD tests from mains power on an RCD board.

Actually perhaps I should have said, wire plug normally ad at socket end park up the E conductor in a connector block (Wago perhaps) and use the L & N normally but put a link in to connect the N & E terminals , therefore your E (on the socket) is not an Earth but really a Neutral !

I can see were you are going with this BUT ......

While this might work as a debugging aid to something that you can put on the test bench & ensure that its has NO connections to anything else, that's so easy with the central heating. The cpc to the heating system is almost certainly connected to the heating system pipework via the boiler/pump & will, in all likelihood, have other connections to the house earth eg. Immersion heater or gas/water bonding etc etc

Your N+E link in your test lead will have effect of creating a second N+E bond in the property, in parallel with the official one in the service head (assuming PME). This could well result in parallel path neutral currents flowing which, at best, would be confusing & at worst may overheat your second N+E link.

 

[ericmark]

I think the idea is only to test the RCD however my RCD tester died a long time ago, so better to take the RCD FCU to work and test it there to borrowing their tester and test it here. It would first need to be included in a test lead. But have no intension of fitting until the summer.

What type of UPS? True sine-wave?

Not a clue, but since the RCD tripped while grid tied can't see that being a problem.

Does it require RCD protection ?

One would hope not, as other wise a compliance certificate should not have been issued. However I watched the heating guy doing the tests with the cover off the boiler, and some cables must be buried within the wall using twin and/or triple and earth flat profile cable. Since it was RCD protected I would have expected it to continue to be RCD protected. The other UPS supplies use SWA and RCD type A sockets. Could not find in local suppliers a type A FCU.

 

[ebee]

This could well result in parallel path neutral currents flowing which, at best, would be confusing & at worst may overheat your second N+E link.

No it would not. It just means that you are testing an RCD without any connection to Earth and you are merely putting a second connection of a N polarity to the Earth connection of the test RCD so you will fool the test meter that an Earth is present, therefore allowing the test to proceed. At no time are any installation conductors connecting an N to an E . The incoming E remains derived from whatever method it is already (TNS, TT, TNC-S) and that remains unchanged

 

[AdrianUK]

No it would not. It just means that you are testing an RCD without any connection to Earth and you are merely putting a second connection of a N polarity to the Earth connection of the test RCD so you will fool the test meter that an Earth is present, therefore allowing the test to proceed. At no time are any installation conductors connecting an N to an E . The incoming E remains derived from whatever method it is already (TNS, TT, TNC-S) and that remains unchanged

I agree this would be true with a an item of equipment in isolation on the test bench.....

BUT with something like Erics central heating boiler .... you cant guarantee the green & yellow wire in the supply cable to the boiler isnt also connected to the installation earth by some other means eg. via an immersion heater?

 

[davelx]

I think the idea is only to test the RCD however my RCD tester died a long time ago, so better to take the RCD FCU to work and test it there to borrowing their tester and test it here. It would first need to be included in a test lead. But have no intension of fitting until the summer.

Not a clue, but since the RCD tripped while grid tied can't see that being a problem.

The UPS is a separate unit from your GTI?
A proper online UPS does double conversion - AC - > DC - > AC so the output waveforms and characteristics are totally separate from the mains input. So whether the input comes from the grid or from your GTI in inverting mode is nether here nor there - it's not the same waveform.

 

[ericmark]

Thank you @davelx that may well explain why it trips. If that is the case, best option is use same socket used by the installers, and fit a plug to central heating supply.